Thursday, June 1, 2006

Keep 'Em All Coming Back

What exactly constitutes loyalty among readers of weblogs? What is loyalty, anyway? Habituated interest? Supersocial attachment? Attentional symbiosis? Casual subscription? And are future returns, tracked through IP addresses, adequate or reliable indications of loyalty, of interest, or of a sort of persistence, of ongoing reawdrite entanglement? Could be.

Seven days ago or thereabouts, I signed up for a trial run of Google Analytics. I was invited, which means that I must've filled out a form some time ago expressing interest in the measures and data presentations, being the graph-geek I am. G.A. runs on a few lines of code; I slipped into the templates for this weblog, just before the </body> tag, as instructed. Within a few days, the reports accumulated enough data to begin suggesting trends and patterns: yet another matrix of activity at this weblog, tracked, I should note, during what's been yet another dull stretch of writing. Rather like a hitless streak: a flat and yawning series of blase-blogging. Still, as if to mock me and my dis-ambition, the numbers twinkle, evoking in me some down-deep fascination with statistics, with counting and with related reports, displays of information.

Google Analytics offers up a number of graphs and charts, ranging from visits and page views to geographic locations and search terms/phrases, and all of the displays are easily re-sorted and layered for different periods of time and for easy-to-see comparisons. I won't be commenting on all of the features here. By no means is this a comprehensive review of G.A. I only want to point, for now, to one of the graphs, "visitor loyalty," because I find it off somehow. I mean that it makes some odd suggestions.

First, notice that the graph cover a one week period, from May 25 to May 31. According to Google, I had a few visits during that stretch. Okay, 1,476. Some differentiation is due. All visits and visitors, as you might expect, are not equal. My best guess is that there are three distinctive groups here: searchers, readers and spammers. Searchers largely make up the left-most counts. That is, they visit once and never return, whether because they're satisfied with what the search summoned (notes on Barthes' photographic image, for example) or because they are not satisfied (recipes for moth poison, let's say). The point is this: searchers are not yet loyal. Their presence is casual, often accidental, ordinarily forgettable, passing and nonchalant.

The third bunch, those 74 visitors who, in one week's time, came by between the frequencies of 9-14 times and 51-100 times are, well, too loyal to be trusted. Rather than be naive and celebrate them as die-hards, I take them to be spa&mers, mostly, although it's conceivable that a few of the 9-14ers could be legitimate readers, I suppose, or folks coming from a lumped-together network ID (e.g., sub-nets on the syr.edu network). Even when the finest, most compelling entries are rolling through EWM on a daily basis, it's unthinkable that this site would draw 51 visits from any reasonable person in a single week. I could start selling hats and t-shirts with that kind of following.

I take the 32 mid-range visitors to be the regulars, those whose loyalty is most in sync with the steadiest currents in this blogstream--a mix of family, colleagues and outliers, some of whom, perhaps, subscribe to one of the RSS feeds. The graph, then, displays this paradox about blogs, audiences and loyalty (even if I'm still reserved about the term as it applies to reading and writing activity): I'm aiming for the belly, for the lowest point in the U-shape between searchers and sp&mmers. I'm not quite sure, but I think that's the point I was writing toward.

Posted by Derek Mueller at June 1, 2006 5:30 PM to Networks